hgoel

5 ms·

1.
by hgoel 2mo ago
Despite being on a site called Hacker News, we seem to often overlook the simple aspect of wanting local AI hardware to hack (not necessarily in the cybersecurity sense) with. I got my local AI hardware because it's an enjoyable hobby
2.
by hgoel 2mo ago
Torrents always seemed like the more sensible way to distribute model weights. Though I have been disappointed that most of these have been spurred on by the misleading claim that abliterated models were being taken down from HuggingFace be
3.
by hgoel 2mo ago
For me the idea came from the discovery of the sites OpenAI's swarms were using to communicate, particularly the detail that one of them ended up being targeted because it allowed writes via GET requests, which OAI's awful sandbox
4.
by hgoel 2mo ago
When I was putting together something similar, I had settled on having a small ring-buffer style storage, say, ~30GB that would be cleared daily or whenever filled. Recording incidents (and humor) is more interesting than actually getting l
5.
by hgoel 2mo ago
Nope, still just humans exploiting humans.
6.
by hgoel 2mo ago
The entire point of Qwen3.8-Next-Flash is to allow inference engines to implement support for the Qwen4 architecture, so they're ready by the time qwen4 is ready for release.
7.
by hgoel 2mo ago
I've experienced the same, IIRC also using NVIDIA's NVFP4 quant. Also just decided to ignore because it didn't seem to cause any real issues. I figure it might be a training thing, since the hallucinated user messages seem to
8.
by hgoel 2mo ago
They've cried wolf too often and hidden too much, absolutely no trust in any of their "reports" anymore.
9.
by hgoel 2mo ago
I end up having the chance to speak on my current research a couple of times a year. Occasionally I'll realize that the audience is some of the most well known names in my field and panic too much to do a good job. Having people ask to
10.
by hgoel 2mo ago
Wasn't it Dario
11.
by hgoel 2mo ago
Also I wonder how this affects the distribution and precision needed for storing the log weights compared to regular ones.
12.
by hgoel 2mo ago
Yeah, I used to think that it'd be fine if the content was pinned off screen but turns out it's worse because I find myself looking away from where I'm walking.
13.
by hgoel 2mo ago
Foveated eye tracking, but also SLAM for the controllers. Also, to be fair, the onboard processing costs less than a separate VR capable PC.
14.
by hgoel 2mo ago
Not to mention that the past few decades have shown that nukes are a major factor in keeping the peace. Conflicts involving nuclear armed countries have been suspended quickly to avoid escalation, while ones involving a party without them h
15.
by hgoel 2mo ago
Do you agree with Sanders' proposal to lock up anyone for 20 years for researching something his proposal doesn't even define?
16.
by hgoel 2mo ago
I don't really get the ick about wearing these kinds of things in public, but, having tried to wear XR glasses in public has shown me that the bigger issue is that these things are at a whole other level of distracting compared even to
17.
by hgoel 2mo ago
That does not reflect with testing I just did locally. When the option of writing and running scripts was available, Qwen3.8-flash did indeed prefer to just do a physics related calculation via code. But, with a fresh chat and tool access t
18.
by hgoel 2mo ago
It's funny we're already at the "actually this isn't very impressive" stage when it was a little over a year ago when we were making fun of LLMs for not being able to add numbers. IC production takes a vast amount o
19.
by hgoel 2mo ago
Plus lower peak power draw
20.
by hgoel 2mo ago
I wonder if this is a result of them trying to cut token consumption by summarizing their RL training data, or maybe it's from how they anonymize user data for training.
21.
by hgoel 2mo ago
I'm baffled that people expressing your opinion don't understand what they're advocating for. Current generation smartphones are able to run AI models, people have managed to run very large models (even if slowly) via streami
22.
by hgoel 2mo ago
A perpetual state of mutually assured destruction is much better for the world than only the US possessing the bomb.
23.
by hgoel 2mo ago
Hmm? I'm running the models locally... 2x Sparks consume ~150W at peak, and they usually spend more time waiting on results from whatever task they're working on, so I imagine that the contribution to my electricity bill is maybe
24.
by hgoel 2mo ago
Well, right now we have ex Anthropic employees telling the media that their terabyte sized models can possibly copy themselves onto the internet and run elsewhere as if the necessary computing resources are ubiquitous. Plus, "we must p
25.
by hgoel 2mo ago
I'd recommend reading the full post :) Specifically: AI ultimately has to live in this reality and face the corresponding limitations. These companies have already consumed much of the world's supply of computing power for the nex
26.
by hgoel 2mo ago
In the real world there aren't any true exponentials, everything eventually saturates as ultimately physics related constraints hit. You can only compress information so much, transfer it so quickly, you can only access resources at a
27.
by hgoel 2mo ago
The HuggingFace incident was the culmination of OAI allowing thousands of agents of various different models - with no clarity on which stages of development they were at (for all we know, some of those models did not have safeguards traine
28.
by hgoel 2mo ago
It's all about wanting to preserve the status quo, complete with all the suffering and strife, because change is scary and uncomfortable. Add in the sincere American belief that everyone else is beneath them, and you get the version wh
29.
by hgoel 2mo ago
Why are we accepting the framing that the LLMs are felony generators, when the only incidences of LLM generated felonies involved misconfigured sandboxes and reckless waste of resources? The companies doing these things without following co
30.
by hgoel 2mo ago
I'm coming up on a month with my 2x GB10 cluster. I was at 1x for a week before I pulled the trigger for the second. Initial setup was a tad annoying because I had to update their firmwares and then power cycle them to get the 200GbE l

More ›